Why some people with mental retardation and/or autism repeatedly and persistently injure themselves, some so severely to the point of tissue damage and often times permanent scarring, has remained a mystery eluding a single solution. Unraveling this mystery poses paradoxial biomedical and behavioral science questions and creates deeply troubling problems for practitioners and family members of affected individuals. Over the past decade, many cases of self-injurious behavior (SIB) have been treated successfully using behavioral interventions that teach communication and other functional skills. Practical problems of implementation, costs associated with long-term treatment, and cases with no clear social profile appearing about 1/3 of the time suggest, however, that there is still much to be learned about why people self-injure. Our overall goals are to improve treatment, refine diagnosis, and to clarify mechanisms underlying different forms of self-injury. Given the severity of self-injury, it is surprising that few of the models have examined in more detail the relation between variables common to SIB and the neurophysiology of pain regulation. The main objective of this project is to evaluate the validity of several of these variables as possible predictors of response to self-injury. Treatments will be based on the hypothesis that some forms of self-injury involve intense stimulation of body sites sufficient to elicit the release and receptor binding of endogenous opioid peptides. Accordingly, treatments will include transcutaneous electric nerve stimulation (TENS)(an opioid agonist treatment) or naltrexone (an opioid antagonist treatment). Predictors will include observationally- based measures of the environmental functions of self-injury, body site location and intensity of self-injury, and salivary baseline levels of three bioactive substances (substance P, metenkephalin, & cortisol). Following initial identification of subjects (age range 4-25) with mold to profound mental retardation and/or autism, our first aim is to observe and describe in detail how frequently self- injury occurs, what its duration and intensity is, and where on the body it is directed. Following this characterization, substance P, met-enkephalin, and cortisol will be noninvasively examined through saliva as markers for altered pain transmission and predictors of response to treatment. After screening and SIB subtyping (i.e., social, nonsocial, or mixed) 37 subjects whose self-injury is primarily nonsocial or mixed will be evaluated over a 16-week period with TENS and the opiate antagonist naltrexone for self-injury. Subjects whose self-injury is primarily socially motivated will be evaluated with TENS and the opiate antagonist naltrexone for self-injury. Subjects whose self-injury is primarily socially motivated will be evaluated with TENS and receive behavioral interventions through a technical assistance service delivery model. Three- and six-month follow-ups will be conducted for each subject.